The market downturn over the last year or so has been a nightmare for many investors. On one end of the spectrum are the “buy and hold” investors who saw their portfolios battered by the broad-based decline in equities. On the opposite end are those who abandoned the equity market completely after taking large losses and who are still sitting on the sidelines. For those needing their money in the next five to seven years, such a move makes perfect sense. However, those with a longer time horizon are probably kicking themselves now that the Wilshire 5000 is up over 40% since its lows of early March. Others are paralyzed by the prospects of a “double-dip” market decline.
Charles Kirkpatrick’s book, “Beat the Market: Invest by Knowing What Stocks to Buy and What Stocks to Sell” (FT Press, 2008), seems to have been written exactly with this scenario in mind. Kirkpatrick believes that the stock market is still the best investment vehicle available, but also thinks it is impossible to predict the market, or the economy. He outlines stockpicking and portfolio management strategies that he believes individual investors can follow to outperform the market while reducing the risk of capital loss.
Kirkpatrick is president of Kirkpatrick & Co., Inc., which specializes in technical research, and he publishes the Kirkpatrick Market Strategist stock advisory newsletter. Kirkpatrick has been a director of the Market Technicians Association (MTA) and holds the Certified Market Technician (CMT) designation. He is also an instructor at the Fort Lewis College School of Business Administration and a two-time recipient of the Charles H. Dow Award from the MTA for excellence in technical research.
Kirkpatrick begins the book by laying out his case as to why he feels investors are better off investing on their own in individual stocks. As the investment industry has become more specialized, investors have been forced to take on a more active role in the investment process. He notes that the shift to “defined-contribution” retirement plans requires us to decide the makeup of our investment portfolios—stocks versus bonds, foreign versus domestic, small-cap versus large-cap, etc. Unfortunately, as Kirkpatrick points out, most individuals are not investment professionals, making the task even more difficult.
For most investors, mutual funds are the investment vehicle of choice, either in retirement plans or in their own personal investment portfolios. However, Kirkpatrick paints a less-than-flattering picture of investment managers, specifically mutual fund managers. He cites a Motley Fool study which found that, between 1963 and 1998, the average mutual fund underperformed the market by 2% a year. [Interestingly enough, however, this has not stopped Motley Fool from starting its own mutual fund.]
Extending these results over 50 years, the study found that $10,000 invested in the average mutual fund would have grown to a respectable $470,000, while $10,000 invested in a market index would have grown to $1.17 million. John Bogle, founder of the Vanguard Funds, concludes from the study: “Our hypothetical fund investor has earned $1,170,000, donated $700,000 to the mutual fund industry, and kept the remaining $470,000.”
Kirkpatrick, who sells his own stock advisory letter, goes on to say that the fee structure of mutual funds negates the incentive for mutual fund managers to act in your best interest. He points out that the fees you pay to own a mutual fund—sales loads, redemption fees, management fees, distribution (12(b)-1) fees, etc.—are not dependent on the performance of the mutual fund; you pay these fees whether the fund rises or falls. Even if a fund is performing poorly, he notes, it can still generate profits as long as new assets are added to the fund pool.
As a result, Kirkpatrick believes that, with only an hour or so of work every week or month selecting individual stocks, investors can free themselves from underperformance while protecting themselves from substantial investment losses. He suggests that an alternative to investing in individual stocks would be to buy low-fee, no-load index funds.
After investing for over 40 years, Kirkpatrick has come to believe that markets trade on facts, the anticipation of new facts, and emotion. As he puts it, the stock market is “the sum of all information known and anticipated, interpretation of that information, and emotional reactions to that information, right or wrong.” However, he points out that individual investors are competing against institutional investors, who are more knowledgeable and better equipped to capture and analyze that information. As a result, Kirkpatrick doesn’t believe it is possible to beat the professionals at their own game. Instead, he advocates developing a mechanical process that depends on indisputable facts to minimize the effects of emotions.
Kirkpatrick believes that investors face several “biases” that have an adverse impact on their financial success: These include impatience, the fear of being wrong, the need for perfection, and a lack of discipline. In order to overcome his own biases, Kirkpatrick has spent a great deal of time testing specific investing methods. Based on these tests, he chose those methods that showed a history of performing well. He cautions against relying on rumors, financial advisors, or investment “gurus” when making investing decisions—except himself, of course.
In “Beat the Market,” Kirkpatrick takes exception to the concept of risk as it is used by investment professionals and academics. In his opinion, popular risk measures are “completely wrong,” as they only take into account the possibility that a stock may fluctuate widely and do not consider capital loss. The result, he notes, is that a stock trading in a straight line carries with it less risk than one that is oscillating. Furthermore, he says a stock that is rising in price can be labeled just as risky as one whose stock price is declining. However, he asks the question: Which would you prefer, a rising or declining stock price? Taking a more practical viewpoint, Kirkpatrick defines risk as the possibility of capital loss. He is more concerned with an investment’s prospects for making money than how widely its price fluctuates.
For Kirkpatrick, drawdown is a more realistic measure of risk. It is the peak-to-trough decline of an investment and measures how much capital loss an investment suffers. For example, if an equity strategy reaches a peak value of $125,000 before reversing course and eventually reaching a low of $95,000 before it starts winning again, drawdown from the peak value to the trough value is 24%. Drawdown doesn’t consider the time it takes to go from peak to trough; it doesn’t matter if this 24% drawdown took place over five weeks, five months, or five years.
Maximum drawdown simply is the largest drawdown experienced over whatever period is being examined. While maximum drawdown can’t tell you the maximum loss a strategy will experience in the future, it does allow you to compare the inherent riskiness of different strategies.
Finally, Kirkpatrick points out that capital loss due to a broad market decline, such as was experienced over the last year, is rarely considered by professionals. When all stocks are declining, Kirkpatrick advocates some “bailout point” below which investors sell all stocks. While he acknowledges the risk of stocks turning around quickly, he feels the small amount you lose offsets the potential of further market declines hurting your capital position. Kirkpatrick notes that most mutual funds try to adhere to some constant ratio of stocks to bonds or cash, which prevents managers from selling off their equity positions during market declines. In Kirkpatrick’s opinion, the “easiest and cleanest” way to avoid significant losses is to go to cash. Table 1 includes a section outlining how Kirkpatrick adjusts his portfolio to account for market risk.
Kirkpatrick reiterates his opinion several times in “Beat the Market” that it is not possible to predict markets. He bases his argument on the fact that professionals with more timely and accurate information fail at the endeavor, so it us unrealistic for individual investors to think they can succeed where the professionals fail. Instead, he feels that a more successful alternative to predicting the markets is a strategy of “reaction.”
Reaction, for Kirkpatrick, means waiting for the market to indicate what it is going to do and then (re)act accordingly. Kirkpatrick reacts when his data—either fundamental or technical—show a pattern that has proven successful in the past. When he sees such a pattern, his reaction follows three steps, which he terms STRACT:
To illustrate the STRACT method, Kirkpatrick offers the following example: Say a study shows that when stocks advance to new 52-week highs, their chances of advancing an additional 10% are 70%. Any stock you are following that nears its 52-week high enters the setup stage. However, at this point, no action is taken as you wait for the trigger—in his example, the stock hitting a new 52-week high. Once the trigger is initiated, your action would be to buy the stock.
Kirkpatrick states that the first investment problems we face are deciding what to buy and how to do so without having to predict anything. He uses three principle methods for selecting stocks:
Kirkpatrick begins his analysis by collecting data from the immediate past. Value and growth investing require accurate fundamental data for each stock in his investing universe. His price strength analysis requires a history of prices for each stock; it measures how a stock price behaves against its immediate past—if a stock is high versus its immediate past, it is said to have price strength.
For his analysis, Kirkpatrick looks at “relative” data—data compared to other data. For example, when he is looking at value, he looks not only at the value of an individual company, but also at its value relative to the value of all other companies. He then attempts to maximize his profits by looking only at those companies with the best value.
Charles Kirkpatrick believes a mechanical approach to investing will help investors avoid their own biases that ultimately cost them money. He further believes that it is impossible to predict market movements. Instead, he follows the STRACT (setup, trigger, and action) technique that helps him react to individual stock movements. His buy and sell triggers are based on “relative” data elements—price-to-sales, reported earnings growth, and price strength. His analysis has led him to three investment models—Growth, Value, and Bargain. By following his buy and sell triggers, Kirkpatrick feels that individual investors can outperform the market by investing in individual stocks with a minimal time commitment.
The only restrictions he imposes are share price and market capitalization minimums. For all three models, Kirkpatrick requires a minimum share price of $10. For the Value Model, he requires a minimum market cap of $500 million, and he uses a $1 billion minimum market cap for the Growth and Bargain Models.
Growth Model
Value Model
Bargain Model
For the Growth Model, Kirkpatrick uses point & figure charts to help in the buy and sell decision process. He only buys stocks for the Growth Model when they are in an upward trend, as indicated by two higher highs in a three-point reversal point & figure chart.
Kirkpatrick does not believe in “buy and hold” investing. Beyond following the sell triggers for individual stocks, he suggests two methods to reduce market capital risk in a portfolio. This first is to select the maximum number of stocks you wish to hold (he mentions 20 as a reasonable number) and then divide this into the amount of money you wish to invest (let’s say $50,000). This means you will invest $2,500 in each of 20 stocks. However, if you can only find 10 stocks in which to invest, you will only put $25,000 into the market and hold the rest in “cash.”
The other method involves monitoring the moving average of your model or hypothetical portfolio (not your actual portfolio). If the current model portfolio declines below its 12-week moving average, Kirkpatrick sells 25% of the portfolio; if the model portfolio value falls below the 26-week moving average, he sells down so that he is only 50% in stocks; if the model portfolio value declines below the 52-week moving average, he sells everything. The process is then reversed for buying back into the market should a downturn force you to move completely to cash.
Growth Model
Value Model
Bargain Model
When looking at a stock’s value, Kirkpatrick uses the price-to-sales ratio, which is the ratio between a stock’s price and the last four quarters of reported sales for the company. He began using the price-to-sales ratio based on James O’Shaughnessy’s analysis in the book “What Works on Wall Street” and because sales data is less likely to be manipulated by company management.
Kirkpatrick’s analysis involved first calculating the weekly price-to-sales ratios for every stock over the period from 1998 to 2006. He then sorted the companies by price-to-sales ratio and then ranked them into percentiles, where the companies with the highest price-to-sales ratios were in the highest percentiles. Once the companies were placed in percentiles based on their valuation, he calculated the relative price performance for each percentile for the proceeding three, six, and 12 months.
What he discovered was that there was an inverse relationship between the relative price-to-sales ratio percentile and the relative performance three and six months forward. In other words, as the price-to-sales percentile increased, the future performance decreased. For periods of 12 months, the relationship between relative valuation and price performance dissipated. Kirkpatrick took his findings to suggest that investors should not focus on periods longer than a year.
Advancing Versus Declining Markets
Kirkpatrick also wanted to see how his various relative values performed in both advancing and declining markets. As we have learned over the last year, the direction of the overall market can play an important role in the performance of individual stocks.
He used a very simple methodology to define advancing and declining markets. He begins by adding together the closing values of the S&P 500 index for each of the last 12 months and creating an average by dividing the total by 12. At the end of the next month, the oldest monthly value is dropped from the total and the latest monthly close is added and again the average value is calculated. When the S&P 500 monthly close is above its 12-month moving average closing price, Kirkpatrick considers the market as advancing; when the monthly close is below the 12-month average closing price, the market is declining.
Returning to his price-to-sales relative analysis, the relationship between the valuation and the future price performance was only about one-third as strong during advancing markets as it was over all market conditions. However, the inverse relationship between valuation and price performance strengthened by about 25% during declining markets. This indicated to Kirkpatrick that relative valuation is a more important selection criterion during a declining market.
When looking at growth factors, Kirkpatrick prefers to look at growth in reported earnings, not predicted earnings. While he admits that reported earnings are not error-free, Kirkpatrick does question the validity and accuracy of forecasted earnings.
In “Beat the Market,” Kirkpatrick describes the way in which he calculates the reported earnings relative rankings for each stock:
“I take the last four quarters of reported earnings for each company and calculate a ratio of this total to the four-quarter total one quarter earlier.”
Kirkpatrick’s goal is to avoid seasonality, which is why he uses reported earnings over a full four quarters. Also, by using operating earnings, he eliminates the impact of special charges or non-recurring items.
He calculates the reported earnings growth for all companies with positive earnings over both four-quarter periods and then ranks them into percentiles, with the companies having the highest growth being in the highest percentile.
As with the study of price-to-sales ratio percentile rankings, Kirkpatrick examined the three-, six-, and 12-month price performance of each percentile. He found, as expected, that there is a positive correlation between earnings growth and subsequent relative price performance. Interestingly, however, the relationship was not as strong as it was for relative price-to-sales. This indicates that reported earnings growth may not be as useful a selection criterion when testing over all market periods. Furthermore, performance turns below-average for stocks with the very highest reported earnings growth. While Kirkpatrick was surprised by these results, they echo the sentiments of investors such as John Neff, who are more interested in strong, but sustainable growth. Companies with high levels of growth cannot be expected to continue at that level over time.
According to Kirkpatrick’s research, the further you go out in time, the weaker the relationship becomes between reported earnings growth and subsequent price performance. This further raises the question in Kirkpatrick’s mind of whether reported earning growth is an effective selection criterion.
As he expected, advancing markets help stocks reporting above-average reported earnings growth. Again, there is a drop-off in performance at the highest levels of growth, but stocks with ultra-high growth still experienced above-average price performance during advancing markets. Kirkpatrick’s research suggests that reported earnings growth is an effective stock selection tool during advancing markets. During bear markets, he found no statistical relationship between reported earnings growth and price performance.
Kirkpatrick’s research indicates that relative price strength is the most reliable short-term stock selection technique. There are a number of ways to calculate relative price strength. Some calculations compare the percentage change in stock price over a defined period to the percentage change in a stock index, such as the S&P 500, over the same period. However, these measures do not necessarily protect you in a down market, as a stock can be falling and still have “strong” relative strength if it is not falling as rapidly as the index.
Kirkpatrick is concerned about capital loss, so his relative strength calculation involves dividing the current weekly closing price by the 26-week moving average of closing prices. He adds up the week-ending closing prices for each of the last 26 weeks and divides this total by 26. For each subsequent week, the oldest price is dropped and the latest weekly close is added to calculate the moving average. He then ranks all the stocks so that those with the highest relative strength are in the highest percentile rank.
Once again, he found that there is a very strong positive relationship between relative price strength and forward price performance. As Kirkpatrick writes in the “Beat the Market” book, “Relative strength seems to breed more relative strength.”
While, over time, this relationship gradually deteriorates, it is still significantly stronger than relative valuation or reported earnings growth. Eventually, moving forward 12 months, the relationship between 26-week relative strength and subsequent price performance all but evaporates.
During an advancing market, Kirkpatrick found that relative price strength should be your primary selection criterion, bettering relative valuation and relative reported earnings growth. This carried over to declining markets as well.
In “Beat the Market,” Kirkpatrick outlines three different selection strategies he has been testing—Growth Model, Value Model, and Bargain Model. Using AAII’s Stock Investor Pro fundamental stock screening and research database, we attempted to replicate these models.
As of August 14, 2009, our database included 9,809 companies. Table 2 lists the specific criteria that subscribers to Stock Investor Pro can use to build the screens. Also, Table 3 lists the custom fields used in the screen process. [Stock Investor Pro subscribers can download a text file of these fields at http://www.aaii.com/files/ci/200909customfields.txt for cutting and pasting into the Custom Field Editor.]
In 1982, Kirkpatrick began testing a hypothetical portfolio of stocks using relative earnings growth, relative price strength, and a chart pattern.
Price Strength
In “Beat the Market,” Kirkpatrick calculates relative strength by dividing the current weekly closing price by the 26-week moving average of closing prices. To attempt to capture the essence of Kirkpatrick’s measure we created a custom field in Stock Investor Pro that is the ratio of the current stock price to the average of the last six monthly closing prices. The numerator will fluctuate with the weekly closing price but the denominator will only change with the end of each month.
| Data Category | Field | Operator | Factor | Compare To |
| Growth List Screen Criteria | ||||
| Custom Fields* | Null Rel Price Strength | Equals | 1 | |
| Custom Fields* | Relative Price Strength | >= | Adjust value until approx. 10% of companies that passed | |
| Null Rel Price Strength criterion pass this filter | ||||
| Custom Fields* | Null Rel Earnings Growth | Equals | 1 | |
| Custom Fields* | Relative Earnings Growth | >= | Adjust value until approx. 10% of companies that passed | |
| Null Rel Earnings Growth criterion pass this filter | ||||
| Price and Share Statistics | Market Cap Q1 | >= | 1000 | |
| Price and Share Statistics | Price | >= | 10 | |
| Value List Screen Criteria | ||||
| % Rank | % Rank-Price/Sales | <= | 30 | |
| Custom Fields* | Null Rel Price Strength | Equals | 1 | |
| Custom Fields* | Relative Price Strength | >= | Adjust value until approx. 10% of companies that passed | |
| Null Rel Price Strength criterion pass this filter | ||||
| Custom Fields* | Null Rel Earnings Growth | Equals | 1 | |
| Custom Fields* | Relative Earnings Growth | >= | Adjust value until approx. 10% of companies that passed | |
| Null Rel Earnings Growth criterion pass this filter | ||||
| Price and Share Statistics | Market Cap Q1 | >= | 500 | |
| Price and Share Statistics | Price | >= | 10 | |
| Bargain List Screen Criteria | ||||
| % Rank | % Rank-Price/Sales | >= | 17 | |
| % Rank | % Rank-Price/Sales | <= | 42 | |
| Custom Fields* | Null Rel Price Strength | Equals | 1 | |
| Custom Fields* | Relative Price Strength | >= | Adjust value until approx. 3% of companies that passed Null | |
| Rel Strength Price criterion pass this filter (relax for more | ||||
| total passing companies) | ||||
| Price and Share Statistics | Market Cap Q1 | >= | 1000 | |
| Price and Share Statistics | Price | >= | 10 | |
| *See Table 3 for information on creating Custom Fields. | ||||
After calculating our own relative price strength field, we are confronted with screening for those stocks in the top 10% of the database (90th percentile or higher). Since we are screening with a custom field we do not have the ability to screen on percentile ranks like we can with many of the pre-built fields in the program. However, a little trial and error enabled us to isolate the top 10%.
As of August 14, 2009, 8,982 of the companies in the database had a valid (non-null) value for the relative price strength field. Setting the relative price strength level at 159% or higher gives us 898 companies; this is the top 10% of the companies with valid price strength data.
Reported Earnings Growth
Kirkpatrick uses a non-standard calculation for earnings growth, which compares operating earnings over the last four fiscal quarters to the four-quarter total in operating earnings one quarter earlier. His goal is to eliminate the impact of seasonality on a company’s earnings.
| Custom Field Name | Formula |
| Relative Earnings Growth | GrowthRate(IIF([Gross operating income Q1] + [Gross operating income Q2] + [Gross operating income Q3] + |
| [Gross operating income Q4] > 0,[Gross operating income Q1] + [Gross operating income Q2] + | |
| [Gross operating income Q3] + [Gross operating income Q4], Null),IIF([Gross operating income Q2] + | |
| [Gross operating income Q3] + [Gross operating income Q4] + [Gross operating income Q5] > | |
| 0,[Gross operating income Q2] + [Gross operating income Q3] + [Gross operating income Q4] + | |
| [Gross operating income Q5], Null), 1) | |
| Null Rel Earnings Growth* | IsFieldNull([Relative Earnings Growth]) |
| Relative Price Strength | ([Price] / (([Price M001] + [Price M002] + [Price M003] + [Price M004] + [Price M005] + [Price M006]) / 6)) * 100 |
| Null Rel Price Strength* | IsFieldNull([Relative Price Strength]) |
|
*The IsFieldNull function will convert a data field into either a 0 value if the field is Null or a 1 value if it is a valid number. Stock Investor Pro subscribers can download a text file of these fields at www.aaii.com/ci/200909/customfields.txt for cutting and pasting into the Custom Field Editor. |
|
Once again we must create a custom field to capture Kirkpatrick’s screening methodology. Our custom field uses operating income, just as Kirkpatrick does, to avoid the special charges or adjustments to earnings. Furthermore, since Kirkpatrick only considers companies with positive earnings, our custom field eliminates those firms with negative operating earnings over either four-quarter period.
As of August 14, 2009, 4,117 companies in the database had non-null relative earnings growth values. A growth rate of 25.9% isolated 411 of those companies—the top 10%.
Price and Market Cap
Lastly, Kirkpatrick looked for growth companies with market capitalizations of at least $1 billion and share prices of at least $10. We are able to recreate both of these filters using Stock Investor Pro.
Passing Companies
The first section of Table 4 lists the five companies passing our Growth List screen as of August 14, 2009. These companies are ranked in descending order by their relative price strength.
Sell Criteria
Once he had “purchased” a stock for his Growth Model, Kirkpatrick followed these rules to determine when to remove a stock from the hypothetical portfolio:
He explained the reason for the chart pattern rule as twofold: First, it confirmed that the stock was actually rising in price and, second, it automatically “sold” a stock if it declined by a predetermined amount. The chart pattern rule served as his stop order. Kirkpatrick determined the previous important support level for the stocks in the portfolio. When a stock in the portfolio broke two of these important support levels, Kirkpatrick removed it from the portfolio.
| Company (Exchange: Ticker) |
Rel Earnings Grth (%) |
EPS Grth 5-Yr (%) |
Relative Price Strgth (%) |
Rel Strgth % Rank 26-Wk (%) |
Price-to- Sales % Rank (%) |
Sales Grth 5-Yr (%) |
|
| Description | |||||||
| Growth List | |||||||
| STEC, Inc. (M: STEC) | 291.7 | -2.3 | 218 | 99 | 87 | 1.4 | memory drives |
| Home Inns & Hotels Management (M: HMIN) | 43.6 | 111.2 | 201.9 | 95 | 78 | na | hotel chain in China |
| SXC Health Solutions Corp. (M: SXCI) | 48.6 | na | 174.9 | 86 | 43 | na | healthcare benefit mgmt |
| Massey Energy Co. (N: MEE) | 114.8 | 29.2 | 166.4 | 83 | 44 | 13.7 | coal producer |
| Ashland Inc. (N: ASH) | 29.8 | 15 | 161.7 | 97 | 19 | 1.3 | specialty chemicals |
| Value List | |||||||
| Clearwater Paper Corp. (N: CLW) | 206.2 | na | 226.6 | 97 | 25 | na | pulp-based prods |
| Ashland Inc. (N: ASH) | 29.8 | 15 | 161.7 | 97 | 19 | 1.3 | specialty chemicals |
| Bargain List | |||||||
| Solutia Inc. (N: SOA) | 23 | 22.3 | 226.5 | 92 | 33 | -2.8 | chemical materials |
| Oshkosh Corp. (N: OSK) | nmf | -1.3 | 209.4 | 95 | 22 | 30 | specialty vehicles |
| Protective Life Corp. (N: PL) | nmf | -17 | 189.1 | 94 | 31 | 5.1 | financial servs holding co |
| Ternium S.A. (ADR) (N: TX) | -51 | 4.3 | 188 | 91 | 40 | 51.6 | invests in steel companies |
| Jones Apparel Group, Inc. (N: JNY) | nmf | -41.2 | 180.1 | 97 | 22 | -3.7 | brand apparel |
| Goodyear Tire & Rubber Co. (N: GT) | nmf | 41.3 | 173.9 | 92 | 17 | 5.2 | manufactures tires |
| Lincoln National Corporation (N: LNC) | nmf | -43.9 | 173.5 | 78 | 39 | 13.3 | insurance & invest mgmt |
| Hertz Global Holdings, Inc. (N: HTZ) | nmf | -49.3 | 173.2 | 88 | 29 | 7.5 | car rental |
| Hartford Financial Services (N: HIG) | nmf | -93.7 | 171.5 | 72 | 28 | -13.2 | insurance & fin’l servs |
| Electrolux AB (ADR) (O: ELUXY) | 10.7 | -8 | 166.7 | na | 23 | -5.2 | home & prof appliances |
| Armstrong World Industries (N: AWI) | -28.6 | 29 | 166.2 | 75 | 30 | 0.8 | flooring products |
| Rockwood Holdings, Inc. (N: ROC) | nmf | -11.6 | 165.5 | 90 | 29 | 33.5 | specialty chemicals |
| W.R. Grace & Co. (N: GRA) | -55.1 | 32 | 163.5 | 90 | 24 | 10.9 | specialty chemicals |
| CBS Corporation (N: CBS) | nmf | -73.3 | 163.1 | 79 | 31 | 0.6 | mass media co |
| International Paper Co. (N: IP) | nmf | -49.9 | 162.7 | 94 | 21 | 2.3 | paper & packaging |
| Ashland Inc. (N: ASH) | 29.8 | 15 | 161.7 | 97 | 19 | 1.3 | specialty chemicals |
| Wyndham Worldwide Corp. (N: WYN) | nmf | -43.3 | 161.5 | 96 | 38 | 10.1 | lodging & vacation rental |
| Swiss Re (ADR) (O: SWCEY) | nmf | -19.9 | 161.3 | na | 34 | -7.3 | reinsurance |
| CB Richard Ellis Group, Inc. (N: CBG) | nmf | -84.1 | 159.4 | 94 | 37 | 25.8 | commercial real estate |
|
Exchange Key: A = American Stock Exchange, M = NASDAQ, N = New York Stock Exchange, O = over the counter. Source: AAII’s Stock Investor Pro/Thomson Reuters. Data as of 8/14/2009. |
|||||||
He assumed that stocks were added to the portfolio in equal dollar amounts, so that the weekly performance of the portfolio was simply the average of the weekly returns for all the stocks in the portfolio. The performance of his hypothetical portfolios did not include transaction costs or dividends. Each week he added any new stocks that matched his selection criteria and eliminated those socks that failed.
Between 1982 and the end of 2007, the Growth Model grew by more than 100, versus 13 times for the S&P 500. Over that time, the hypothetical portfolio turned itself over almost twice a year. Kirkpatrick uses this Growth Model to generate his Growth List, which he publishes in his weekly newsletter.
Kirkpatrick’s Growth List combines quantitative filters for relative price strength and relative reported earnings growth, and then requires point & figure chart analysis to determine whether the stock is in an upward trend.
Despite the success of his Growth Model, Kirkpatrick was concerned about the fact that its performance had occurred during one of the strongest bull markets in history. He wanted to strengthen the system against capital loss to protect against the inevitable market reversal. He believed relative price strength would not be effective during a market downturn and could lead to significant capital losses. For Kirkpatrick, the alternative was to reduce the risk of the portfolio by beginning with a group of stocks with low valuations. His reasoning was that, since the low valuations probably were due to a decline in price, the downside risk had been reduced. Using relative price-to-sales percentiles, Kirkpatrick arbitrarily selected only those stocks in the 30th percentile or lower.
He tested his Value Model from 1998 to 2007, where it outperformed the Growth Model and the S&P 500 index.
Kirkpatrick’s Value List mimics the Growth List, except that it uses the relative price-to-sales ratio as a means of reducing risk instead of using a chart pattern stop.
To begin our Value List screen, we use the same screening criteria we used for the Growth List screen regarding relative price strength and relative earnings growth.
Relative Valuation
For his Value List, Kirkpatrick chooses stocks with relative price-to-sales ratios that are in the 30th percentile or lower. In Stock Investor Pro, price-to-sales is one of the fields that you can screen upon using percentile rank. Therefore, for our Value List screen, stocks must have a price-to-sales percentile rank than is less than or equal to 30.
Price and Market Cap
For the Value List, Kirkpatrick relaxed his market cap requirements to include companies whose market capitalization is $500 million or higher. He did, however, maintain the $10 share price minimum. We are able to recreate both of these filters using Stock Investor Pro.
Passing Companies
The second section of Table 4 shows the two companies passing our Value List screen as of August 14, 2009. These companies are ranked in descending order by their relative price strength.
Sell Criteria
Once he had purchased a stock for his Value List, these are the rules Kirkpatrick followed to determine when to remove a stock from the portfolio:
Between January 2005 and December 2007 Kirkpatrick tested a new model, called the Bargain Model, using the best triggers found in his testing of relative value, relative reported earnings growth, and relative price strength outlined earlier.
In 2007, the Bargain Model gained 77.3%, versus a gain of 25.3% for the Value Model and a 43.0% gain for the Growth Model. However, Kirkpatrick admits that several more years of testing are needed before labeling the Bargain Model a successful stock selection methodology.
Price-to-Sales
While the Value List selects stocks with relative price-to-sales in the 30th percentile or lower, Kirkpatrick’s testing of relative price-to-sales ratio percentile rankings indicated optimal performance in percentiles greater than 17 but not higher than the 42nd percentile. Relative valuations outside this range tended to underperform the market.
Therefore, for our own Bargain List screen, we required companies to have a price-to-sales percent rank that is greater than or equal to 17 and less than or equal to 42.
Reported Earnings
Kirkpatrick’s analysis of earnings growth and future price performance indicated only a weak correlation between the two. As a result, he decided to omit relative earnings growth as a selection criterion for the Bargain List.
Relative Price Strength
For both the Growth List and Value List, Kirkpatrick had been selecting stocks with relative price strength in the 90th percentile or higher. When he started the Bargain List, setting the bar at the 90th percentile for relative price strength resulted in too many passing companies to manage in a portfolio. To reduce the number of passing companies, Kirkpatrick upped the requirement to only include companies in the 97th percentile or higher.
Again we must manually adjust the relative strength screening variable to arrive at the desired number of passing companies. However, this time we are looking to isolate only 3% of the 8,982 of the companies in the database with a valid (non-null) value for the relative price strength field, or approximately 269 companies. Setting the relative price strength level at 219% results in 269 companies passing the filter.
However, adding this filter to the rest of the criteria Kirkpatrick uses for the Bargain List results in only one passing company. Using the 159% minimum relative price strength value from the Growth List and Value List screens to capture roughly the top 10% of companies based on relative price strength nets us 19 passing companies.
Price and Market Cap
Lastly, Kirkpatrick looked for Bargain List companies with market capitalizations of at least $1 billion and share prices of at least $10. We are able to recreate both of these filters using Stock Investor Pro for our Bargain List screen.
Passing Companies
The last section of Table 4 lists the 19 companies passing our “relaxed” Bargain List screen as of August 14, 2009 (relative price strength percentile of 90 or higher instead of 97 or higher). These companies are ranked in descending order by their relative price strength.
Sell Criteria
Once he has purchased a stock for his Bargain List, these are the rules Kirkpatrick follows to determine when to remove a stock from the portfolio:
Charles Kirkpatrick takes an extremely mechanical approach to stockpicking. To him, stocks are merely symbols. He does not concern himself with what the company does. He merely allows his data analysis to dictate when to buy and sell stocks. By basing his selection process on relative variables that have tested very well over an extended period of time, Kirkpatrick believes he has found strategies that will perform well in both bull and bear markets and will alert you to sell in time to avoid large capital losses.
I am reading (and posting) this in January of 2012 - nearly three years after the column was written. But it took me until about halfway through the piece to realize that, and put it in a temporal context.
I really wish AAII would briefly note in their "Computerized Investing" emails that they are linking you to an old article because they have nothing new to send along. Many or most of the principles may still be useful, but it's jarring and confusing to have something presented as implied "new" when it is actually rehashed from the previous decade.
posted about 1 year ago by Mark from Maryland
Mark-- I agree with you. I have noticed the same thing. I think if there are no new articles, they should just say so. Recycling the old ones is not cool unless maybe they add a section called "Old Articles" or something so I can maybe skip it. Is AAII reading these comments?
posted about 1 year ago by Jw from Utah
Thank you for your comments. The original publication date of each article we link to is provided at the top of the article. In this case, September 2009. As Mark noted, we link to old articles because the concepts themselves are timeless. We also do this to make readers aware of the deep archive of articles they have at their disposal. Wayne A. Thorp, CFA, editor, Computerized Investing.
posted about 1 year ago by Wayne from Illinois
Wayne: FOr me, it would add some value if AAII added some updated information. e.g., a simple Editor's note or other reference to a recent Mark Hulbert or similar study that proves or disproves or otherwise addressess the article's thesis. This article screams for a Mark Hulbert type validation - yet the original article and this republished version do not mention such outside objective conclusions.
posted about 1 year ago by James from Washington
LIke the other comments I feel that the article date should appear in the email prior to hitting the link where the article is shown. The email says nothing about when the piece was written.
This email is just a re-run of a re-run which could be certainly be informative but to me is frustrating when I thought I was accessing updated material. I didn't realize it until I was half way through it.
posted 11 months ago by John from New York
Sorry, you cannot add comments while on a mobile device or while printing.